Collagen peptide supplementation in combination with resistance training improves body composition and increases muscle strength in elderly sarcopenic men: a randomised controlled trial

The researchers tested whether adding a daily collagen peptide supplement (15g) to a standardised resistance training programme would improve body composition and muscle strength more than resistance training plus a placebo (silica) in elderly men who already had age-related muscle loss (sarcopenia).

**Intervention:** 15g of collagen peptides (hydrolysed type I collagen, mean molecular weight ~3 kDa, brand BODYBALANCE™ by GELITA AG) dissolved in 250ml water, consumed within 1 hour after each training session, and at a similar time on non-training days.

**Comparator:** 15g of silicon dioxide (silica, brand Sipernat 350) dissolved in 250ml water, consumed on the same schedule. Silica was chosen because it is a safe food additive with negligible intestinal absorption and no metabolic effects.

**Primary outcome:** Change in fat-free mass (FFM) measured by dual-energy X-ray absorptiometry (DXA).

**Secondary outcomes:** Change in fat mass (FM), bone mass (BM), isokinetic quadriceps strength (IQS) of the right leg, and sensory motor control (SMC) measured by a one-leg stabilisation test.

**Sample size:** 53 men completed the study (26 in the collagen group, 27 in the placebo group). 60 were randomised; 7 dropped out due to non-compliance (missing >10% of training sessions).

**Population:** Healthy men aged >65 years (mean age 72.2 years, SD 4.68) with sarcopenia class I or II, recruited via newspaper advertisement in Freiburg, Germany.

**Inclusion criteria:** Self-reported considerable loss in muscular strength or physical performance within the last 3–4 years; able to participate in a 3-month resistance training programme; free of acute diseases or illness-related cachexia.

**Exclusion criteria:** Chronic illnesses (liver, kidney, cancer without recurrence for 5 years, cardiovascular disease, advanced arthrosis) or other diseases making exercise participation impossible; absence of sarcopenia confirmed by DXA.

**Sarcopenia definition:** Handgrip strength <32 kg (screened with handheld dynamometer), then confirmed by DXA muscle mass. Class I = muscle mass 1–2 SD below young adult mean; Class II = muscle mass >2 SD below young adult mean.

**Body composition:** Dual-energy X-ray absorptiometry (DXA) using a Stratos DR 2D Fan Beam scanner (Degen Medizintechnik). Measured fat-free mass (FFM), fat mass (FM), and bone mass (BM) before and after the 12-week intervention.

**Muscle strength:** Isokinetic quadriceps strength (IQS) of the right leg measured using a Con-Trex dynamometer. This measures torque (in Newton-metres, Nm) during a controlled-speed leg extension movement.

**Sensory motor control (SMC):** Standardised one-leg stabilisation test using a Posturomed device (Haider-Bioswing), which measures postural sway and balance control.

**Dietary intake:** 4-day nutritional protocols (using household measurements) completed before and at the end of the study, analysed using PRODI 6.0 software.

**Compliance:** Checked by collecting unused supplements and daily records of timing of ingestion, side-effects, or problems.

**Safety:** Blood samples collected at baseline and end of study (ESR, haemogram, creatinine, creatine kinase, urea, ALT, AST).

**Study design:** Randomised, double-blind, placebo-controlled trial.

**Randomisation:** Subjects were randomly assigned to treatment (collagen) or placebo group using a random number generator. The randomisation sequence was concealed from investigators and participants.

**Blinding:** Double-blind — neither the investigators nor the participants knew group assignments until all data were entered, the dataset was secured, and statistical analysis was performed. The placebo (silica) was matched in appearance and preparation (powder dissolved in 250ml water).

**Duration:** 12 weeks of intervention, with measurements taken at baseline and immediately after the 12-week period.

**Exercise protocol:** Supervised resistance training three times per week, 60 minutes per session, at the University of Freiburg. Exercises included pull down, leg press, bench press, back press, and other exercises involving all major muscle groups. Intensity was progressively increased: weeks 1–4: 15 repetitions; weeks 5–9: 10 repetitions; weeks 10–12: 8 repetitions; all at 4 seconds per repetition. Individual adaptations were made based on actual performance. Subjects missing >10% of training sessions were excluded.

**Supplement timing:** Collagen or placebo was consumed within 1 hour after each training session. On non-training days, it was consumed at a similar time of day. No other food was allowed during the first hour after training (except water).

**Statistical approach:** Normality tested with Kolmogorov-Smirnov test. Baseline differences tested with unpaired t-test. Within-group changes tested with paired t-test. Between-group differences tested with two-way repeated-measures ANOVA (factors: group and time). Data presented as means with standard deviations.

**What this design can prove:** This design can establish causality — if the collagen group shows greater improvements than the placebo group, it can be attributed to the supplement rather than to the training alone or to placebo effects. The double-blind design eliminates expectation bias from both participants and researchers. The randomised allocation controls for confounding variables (age, baseline fitness, diet) being unevenly distributed between groups.

**What this design cannot prove:** It cannot determine whether collagen would work without resistance training (no exercise-only control group). It cannot determine the optimal dose (only 15g was tested). It cannot determine whether effects persist after supplementation stops (no follow-up period). It cannot determine mechanisms (e.g., whether effects are due to collagen's amino acid profile, its effect on connective tissue, or other pathways). It cannot generalise to women, younger adults, or non-sarcopenic populations.

**Methodological strengths:** Double-blind design; objective DXA measurements; supervised, standardised training; compliance monitoring; exclusion of subjects with poor adherence; silica placebo (metabolically inert).

**Methodological weaknesses:** Relatively small sample (53 completers); no intention-to-treat analysis reported (7 dropouts excluded); no power calculation reported; no adjustment for multiple comparisons; industry funding (collagen provided by GELITA AG, a collagen manufacturer); short duration (12 weeks); no long-term follow-up; only men studied; dietary intake assessed only by self-report (4-day food diaries); no measurement of habitual protein intake during the intervention (only at baseline and end); no assessment of blinding success (whether participants could guess their group).

**Primary outcome — Fat-free mass (FFM):**

Collagen group: increased by +4.2 kg (SD 2.31) from baseline

Placebo group: increased by +2.9 kg (SD 1.84) from baseline

Between-group difference: +1.3 kg favouring collagen (P < 0.05)

Both groups showed significant increases (P < 0.01 within groups)

**Secondary outcomes:**

**Fat mass (FM):**

Collagen group: decreased by -5.4 kg (SD 3.17)

Placebo group: decreased by -3.5 kg (SD 2.16)

Between-group difference: -1.9 kg favouring collagen (P < 0.05)

Both groups showed significant decreases (P < 0.01 within groups)

**Isokinetic quadriceps strength (IQS):**

Collagen group: increased by +16.5 Nm (SD 12.9)

Placebo group: increased by +7.3 Nm (SD 13.2)

Between-group difference: +9.2 Nm favouring collagen (P < 0.05)

Both groups showed significant increases (P < 0.01 within groups)

**Bone mass (BM):**

Both groups showed significant increases (P < 0.01 within groups)

No significant between-group difference reported (collagen not superior to placebo for bone mass)

**Sensory motor control (SMC):**

Both groups showed significant improvements (P < 0.01 within groups)

No significant between-group difference reported (collagen not superior to placebo for balance)

**Dietary intake:**

No significant differences in energy or macronutrient intake between groups at baseline or after intervention

This rules out the possibility that one group simply ate more protein overall

**Compliance and safety:**

No adverse events reported

Blood markers remained within normal ranges

Compliance was checked by collecting unused supplements (exact compliance rate not reported numerically)

**Fat-free mass:** The collagen group gained an extra 1.3 kg of lean mass beyond what training alone produced. To put this in perspective, this is roughly equivalent to the lean mass of a small chicken breast. Over 12 weeks, this represents about 0.11 kg (110g) of additional lean mass gained per week — a modest but meaningful difference for elderly men who are losing muscle at a rate of 1–2% per year.

**Fat mass:** The collagen group lost an extra 1.9 kg of fat beyond the training-only group. This is roughly equivalent to the fat content of about 17,000 calories — meaning the collagen group effectively "burned off" an additional 200 calories per day compared to the placebo group, despite no differences in reported dietary intake.

**Muscle strength:** The collagen group gained an extra 9.2 Nm of quadriceps torque. To make this tangible: a typical leg extension exercise at the gym might involve 50–100 Nm of torque for an elderly man. An additional 9.2 Nm represents roughly a 10–15% improvement in strength beyond training alone — enough to potentially make daily activities like standing from a chair or climbing stairs noticeably easier.

**Relative improvement:** The collagen group's FFM gain was about 45% greater than the placebo group (+4.2 kg vs +2.9 kg). Their strength gain was more than double (+16.5 Nm vs +7.3 Nm). Their fat loss was about 54% greater (-5.4 kg vs -3.5 kg).

**What the authors acknowledge:**

The study was conducted only in men, so results may not apply to women

The sample was relatively small (53 completers)

The duration was only 12 weeks, so long-term effects are unknown

The specific mechanism of action was not investigated

**What a critical reader would note:**

**Industry funding:** The collagen supplement was provided by GELITA AG, a company that manufactures and sells collagen products. The Collagen Research Institute GmbH (co-author affiliation) also has a clear interest in positive results. While the study was double-blind and randomised, industry-funded nutrition studies are more likely to report positive findings.

**No intention-to-treat analysis:** 7 of 60 randomised subjects (11.7%) dropped out and were excluded from analysis. If dropouts differed between groups or had different outcomes, this could bias results.

**No correction for multiple comparisons:** The study tested multiple outcomes (FFM, FM, BM, IQS, SMC) but did not adjust P-values for multiple testing. With 5 outcomes, the chance of at least one false positive is ~23%.

**Silica placebo:** While metabolically inert, silica is not a protein. This means the study compared "protein + training" vs "no protein + training," not collagen specifically vs another protein. The results could simply reflect that adding any protein (15g) to training is beneficial.

**No protein-matched control:** The placebo group did not receive an isonitrogenous (equal protein) placebo. So the study cannot distinguish between "collagen is uniquely effective" and "any extra 15g of protein is beneficial."

**Self-reported diet:** Dietary intake was assessed by 4-day food diaries, which are notoriously inaccurate (people underreport calories by 20–50%). If collagen supplementation changed appetite or food choices, this might not have been captured.

**No measurement of habitual protein intake during the intervention:** Diet was only assessed at baseline and end, not throughout the 12 weeks. Changes in protein intake from other sources could have confounded results.

**No blinding assessment:** The authors did not report whether participants could guess which group they were in. Collagen has a distinct taste and texture; silica may have been distinguishable.

**Population specificity:** Results apply only to sarcopenic elderly men (mean age 72). Younger adults, women, or healthy older adults without sarcopenia might respond differently.

**No follow-up:** We don't know if gains were maintained after stopping supplementation or training.

For someone running their own n=1 experiment:

**What to test:**

15g of hydrolysed collagen peptides (type I, ~3 kDa molecular weight, such as BODYBALANCE™ or similar products) taken daily

Combined with a progressive resistance training programme (3x/week, targeting all major muscle groups)

Consume within 1 hour after training; on rest days, take at a consistent time

**Minimum meaningful duration:**

12 weeks (the study duration). Some effects may appear earlier, but the full benefit likely requires at least 8–12 weeks

Measure at baseline, week 6, and week 12 to track trajectory

**What to measure (specific metrics):**

**Body composition:** Use DXA if available (gold standard). Alternatively, use bioelectrical impedance analysis (BIA) or skinfold calipers for fat-free mass estimates. Measure at the same time of day, same hydration status, same time since last meal.

**Muscle strength:** Measure 1-rep max (1RM) on leg press or bench press, OR use a handheld dynamometer for grip strength, OR measure isometric knee extension force. Test at the same time of day, after a standardised warm-up.

**Fat mass:** Waist circumference (measured at navel level, after exhale) and body weight (morning, after voiding, before eating).

**Dietary control:** Log all food intake using an app (e.g., MyFitnessPal, Cronometer) to ensure total protein intake is consistent between phases. Aim for at least 1.2–1.6g protein/kg bodyweight/day total.

**Subjective measures:** Rate of perceived recovery (1–10 scale), muscle soreness (1–10 scale, 24h post-training), and ability to perform daily activities (e.g., chair stand test, stair climbing).

**Key confounds to control for:**

**Total protein intake:** If you increase collagen by 15g, you must keep total daily protein constant between test and control periods. Otherwise, you're testing "more protein" not "collagen specifically." Reduce other protein sources by 15g when taking collagen.

**Training consistency:** Keep training volume, intensity, and frequency identical between phases. Use a training log to track sets, reps, and weights.

**Timing:** Take collagen at the same time relative to training (within 1 hour post-exercise) consistently.

**Dietary consistency:** Keep total calories, macronutrient ratios, and meal timing consistent between phases. Use a 7-day meal plan repeated in each phase.

**Sleep and stress:** Track sleep quality (hours, subjective quality) and daily stress (1–10 scale) as these affect muscle protein synthesis.

**Hydration and measurement timing:** Always measure body composition at the same time of day, same hydration status (e.g., first thing in morning after voiding).

**Supplement purity:** Use a single batch of collagen from a reputable manufacturer that provides third-party testing for heavy metals and contaminants.

**What a positive result